BACKGROUND OF THE INVENTION
Field of the Invention
[0001] The present invention is related to a method of isolating an antibody as defined
in the appended claims.
Background Art
[0002] Biological macromolecules (
i.e., biomacromolecules) such as recombinant biomacromolecules have importance in a diverse
array of technologies. Traditionally, biomacromolecules have been purified using many
different methods,
e.
g., filtration, centrifugation, size exclusion chromatography, affinity chromatography,
and combinations of the above, just to name a few. The method of purification is generally
chosen based on a characteristic of the biomacromolecule that distinguishes it from
one or more impurities that coexist with the biomacromolecule in a composition. A
vast number of biomacromolecules are commercially important, and an ability to purify
a large amount of biomacromolecules in a timely and cost effective manner is desired.
Extensive research has been performed to increase efficiency of current purification
technologies and methods for purifying biomacromolecules. Often, purification techniques
that are suitable for small scale preparations are not suitable for industrial-scale
purification.
[0003] Commercially important biomacromolecules include, e.g., proteins and nucleic acids,
e.g., DNA and RNA. Two examples of biomacromolecules that are often isolated on an industrial
scale are monoclonal antibodies and fusion proteins. These antibodies and fusion proteins
are valuable in various diagnostic and therapeutic fields, and have been used to treat
various diseases such as inherited and acquired immune-deficiency diseases and infectious
diseases.
[0004] Traditional approaches to producing purified antibodies include ammonium sulfate
precipitation, use of caprylic acid followed by centrifugation, ion exchange chromatography
(e.g., DEAE or hydroxyapatite), immunoaffinity purification
(e.g., protein A or protein G), and dialysis. See
e.g., Antibodies: A Laboratory Manual, Harlow and Lane, Cold Spring Harbor Laboratory (1988). The use of a combination of the above methods is common, e.g., antibody purification
from plasma using ethanol fractionation followed by ion exchange chromatography and/or
caprylic acid (CA) precipitation. See for example
McKinney et al., J. Immunol. Methods 96:271-278 (1987);
U.S. Pat. Nos. 4,164,495;
4,177,188;
RE 31,268;
4,939,176; and
5,164,487. In addition, acidification of fermentation has been used to improve recovery and
stability of antibodies and recombinant proteins. See
e.g., Lydersen et al., Annals New York Academy of Sciences 745:222-31 (1994).
[0005] Various other methods have been developed for isolation and/or purification of antibodies
including the application of acid precipitation. See
e.g., U.S. Pat. Nos. 7,038,017;
7,064,191;
6,846,410;
5,429,746;
5,151,504;
5,110,913;
4,933,435;
4,841,024;
4,801,687. and
WO 2006/012500. However, many of these methods can result in large feedstock volumes and recovery
loss and/or have a high cost for producing antibodies on industrial scales. Limited
work has exploited the benefits of adjusting harvest conditions to improve cellular
clarification robustness, especially as it pertains to tangential flow filtration.
[0006] The harvesting of antibodies and recombinant proteins from industrial-scale bioreactors
containing mammalian or bacterial cells is generally performed using either filtration
or centrifugation. However, in the case of these techniques, nucleic acid (
e.g., DNA), host cell proteins (HCP), and growth media components are often not adequately
separated from the biomacromolecule of interest. The recent drive for generating increased
amounts of protein production in cell culture has required bioreactors to operate
at higher cell densities, which increases the amount of impurities such as DNA, HCP,
and other media components. The elevated levels of contaminants have placed stronger
demands on both cell harvesting operations (e.g., the filtration and centrifugations
steps), as well as the downstream purification steps (e.g., chromatography and dialysis
steps). The addition of these higher levels of impurities may increase the number
of purification steps that need to be performed, thus decreasing overall production
throughput.
[0007] As a result of the aforementioned difficulties and inefficiencies, there is a need
to improve the strategy for isolation of biomacromolecules.
BRIEF SUMMARY OF THE INVENTION
[0008] The present invention is directed to a method an antibody according to the appended
claims.
BRIEF DESCRIPTION OF THE DRAWINGS/FIGURES
[0009]
FIG. 1 shows analysis of particle size distribution for a monoclonal antibody (Mab)
cell culture at a pH of 4.5, 5.75 and 7.1 (unadjusted). At pH 4.5, there is a pronounced
decrease in the proportion of particles with less than a 2.5 µm diameter, and an increased
population of particles from 2.5 to 6.0 µm.
FIG. 2 represents magnified images of cells at a pH of 4.5 (FIG. 2c), 5.75 (FIG. 2b)
and 7.1 (FIG. 2a), stained with Trypan Blue. The experimental images indicated increased
staining at lower pH levels.
FIG. 3 shows supernatant turbidity in cell culture as a function of a decrease in
pH. Generally, the turbidity is reduced at lower pH levels due to cellular and particle
flocculation.
FIG. 4 represents the effect of pH adjustment of the harvest feed on turbidity of
the supernatant. The y-axis represents the turbidity of the supernatant after harvest
pH adjustment and settling of flocculated cells + cell debris. The x-axis represents
the various pH adjustment values of the harvest material. In particular, the data
was generated by pH adjusting aliquots of harvest feed to the specified pH using 25%
v/v acetic or citric acid, allowing flocculation and settling of the cellular mass
to occur, and measuring the turbidity (degree of clarity) of the clear supernatant.
The data in FIG. 4 is generated from various harvest streams containing different
recombinant proteins including antibodies and fusion proteins. The graph demonstrates
a general reduction on supernatant turbidity as the pH of the adjusted harvest material
drops, indicating a clearer supernatant at lower pH value. This is a result of higher
degree of cellular flocculation that occurs at lower pH resulting in more rapid settling
of the cellular mass. This flocculation and improved settling provides for improved
mass transfer near membrane surface and thus more efficient microfiltration operational
performance (See FIG. 12 and FIG. 13).
FIG. 5 represents the effect of various pH levels on removal of host cell proteins
(FIG. 5a) and DNA (FIG. 5b) from two separate cell culture, the cell cultures producing
antibodies A and B. The x-axis represents the various pH adjustment values of the
cell culture. The y-axis represents the % of either host cell protein or DNA remaining
in the sample.
FIG. 6 represents the effect of various pH levels on product quality and recovery
for two different monoclonal antibodies, A (FIG. 6a) and B (FIG. 6b). The x-axis represents
the various pH adjustment values of the cell culture. For the MabA graph (FIG. 6a),
the squares represent protein recovery, the circles represent SEC monomers, and the
triangles represent IRC main isoforms. For the Mab B graph (FIG. 6b), the diamonds
represent protein recovery, the Xs represent SEC monomers, and the circles represent
IEC main isoforms.
FIG. 7 represents the effect of various pH levels on microfiltration performance.
The x-axis represents the various pH adjustment values of the cell culture. The Xs
represent concentration percentages of host cell proteins, the stars represent concentrations
of DNA, the dark triangles represent turbidity, the light triangles represent SEC
monomers, and the astericks represent IEC main isoforms.
FIG. 8 represents the effects of various pH levels on the yield of monoclonal antibody
A after microfiltration. The measurements at pH levels greater than 5.5 failed due
to fouling of the membrane.
FIG. 9 represents the effects of various pH levels on the isolating monoclonal antibody
B after centrifugation. The diamonds represent recovery of antibody B, the squares
represents host cell protein concentrations, the triangles represent DNA concentrations,
the plus signs (+) represent turbidity measurements, the circles (O) represent SEC
monomers, and the Xs represent IEC main isoforms.
FIG. 10 shows a schematic of a microfiltration system. The objective of this system
operation is to remove cells and other debris from the cell culture in advance of
chromatographic purification. The clarification is performed using microfilters (0.2
- 0.65µm nominal pore size) in a crossflow configuration (harvest feed and retentate
stream flow parallel to the membrane surface). This system includes a peristaltic
pump to circulate the feed through the microfiltration (MF) cartridges; however, other
pumps can be used such as a rotary lobe or diaphragm pump. During filtration, the
elapsed time, permeate stream weight, and the cartridge inlet, outlet, and permeate
stream pressures are monitored. After the cell culture conditioned medium is concentrated
5-7-fold, a constant volume diafiltration with a buffered solution (pH similar to
the harvest feed pH) is performed to recover most of the remaining product. The permeate
stream is drawn from the system at a constant flow rate and collected in a vessel.
The temperature of the operation can occur between 2-26°C. The feed vessel contains
either unadjusted or pH adjusted harvest feed and is agitated to prevent cell settling.
FIG. 11 shows the method for calculating the transmembrane pressure drop (TMP) across
the microfilter. The TMP is calculated from the pressure measured on the harvest feed,
retentate stream, and permeate stream sides. The pressure measurement is performed
using a needle pressure gauge, a digital gauge, or a pressure transducer.
FIG. 12 represents the effect of pH adjustment on microfiltration operational performance.
The y-axis represents the microfiltration filter average transmembrane pressure difference
(TMP) in psi using 0.65µm pore size hollow fiber filters; the x-axis represents the
ratio of the volume of harvest feed loaded on the membrane to the area of the membrane
(L/m2). The harvest material used for these microfiltration experiments comprises Chinese
Hamster Ovary cells that produce a glycosylated humanized monoclonal antibody (IgG1).
The legend of Fig. 12 shows data for various pH adjusted harvest streams, and various
percent cell viabilities. The graph shows that filtering unadjusted harvest feed which
is lowered from neutral Bioreactor conditions to a pH of 4.7-5.3 results in a lower
overall transmembrane pressure across the filters relative to filtering harvest feed
which remains at a pH ranging between 6.8 and 7.2. The data indicates that lowering
the pH of the harvest streams allows for reduced fouling of the filters at a set permeate
flux (as indicated by the lower TMP at high loading), resulting in a more robust microfiltration
operation.
FIG. 13 represents the effect of pH adjustment on microfiltration operational performance
similar to FIG. 12 but for another monocolonal antibody. The y-axis represents the
microfiltration filter average transmembrane pressure difference (TMP) in psi using
0.65µm pore size hollow fiber filters; the x-axis represents the ratio of the volume
of harvest feed loaded on the membrane to the area of the membrane (L/m2). The harvest material used for these microfiltration experiments comprises Chinese
Hamster Ovary Cells that produce a well differentiated humanized, monoclonal antibody.
The FIG. 13 legend shows data for various pH adjusted harvest streams, and various
percent Cell Viabilities. The graph shows that filtering harvest feed which is lowered
from neutral Bioreactor conditions to a pH of 4.7-5.2 results in a lower overall transmembrane
pressure across the filters relative to filtering unadjusted harvest feed which remains
at a pH ranging between 6.8 and 7.2. The data indicates that lowering the pH of the
harvest streams allows for reduced fouling of the filters at a set Permeate flux (as
indicated by the lower TMP at high loading). This results in a more robust microfiltration
operation.
FIG. 14 represents the effects of pH change and the presence of different divalent
cations on the product protein titer in the conditioned harvest stream The y-axis
represents the product titer, normalized to the titer of pH 7.0 harvest which has
no cations added. The x-axis represents the harvest streams for various biomacromolecules
at several pH values (pH 7, 5, or 4) and types of divalent cations. The data for the
fusion protein demonstrates that addition of Co2+ during the pH 5 adjustment preserves the fusion protein in solution or eliminates
possible protein co-precipitation with the flocculated cells and cell debris. The
data for the antibodies demonstrates that addition of Mg2+ or Ca2+ during the pH 5 adjustment preserves the antibodies in solution or eliminates possible
co-precipitation with the flocculated cells and cell debris.
FIG. 15 represents the effect of harvest pH adjustment on DNA removal for different
harvest streams containing various antibodies. The y-axis shows the amount of DNA
impurity present (µg) per amount of antibody present (kg) in the clarified harvest
stream. The x-axis represents the harvest/antibody stream both unadjusted and pH adjusted
to pH 4.7 with 25% acetic acid. For all antibodies shown below, the adjustment of
harvest pH from about 7.0 to 4.7 effectively brings about precipitation of DNA, resulting
in a clarified harvest stream having 1.5 to 3 logs reduction in DNA impurity levels.
FIG. 16 represents the effect of adding various cations to the harvest feed followed
by pH 5 adjustment of the harvest. The y-axis shows the loss in % protein based on
the unadjusted harvest stream with or without the presence of additional ions. The
first bar on the left of the graph represents protein loss of 14% (due to precipitation)
after pH harvest adjustment to pH 5.0 without any additional divalent ions. The remaining
bars represent additional protein loss (>14%) or protein product regain (< 14%) in
the presence of various divalent cations. The data shows that the presence of various
transition metal ions such as Ni2+, Ca2+, Mg2+, Mn2+, and Co2+ improve protein recovery.
FIG. 17 represents the effect of CoCl2 concentration on protein titer recovery from a harvest stream that is pH adjusted
to 5.0 with 25% acetic acid. The y-axis represents the loss in % protein of the pH
adjusted harvest stream at various levels of CoCl2. The x-axis represents the concentration of CoCl2 (mM) present in the pH 5.0 adjusted harvest stream. The data show that as the concentration
of Co2+ ions increases, the % loss of protein product due to pH induced precipitation decreases.
An equilibrium concentration of 10mM is shown for the fusion protein.
FIG. 18 represents the effect of the method of the present invention on filter protein
rejection and overall clarification recovery. The y-axis represents the instantaneous
filter protein rejection defined by the retention coefficient (R). The x-axis represents
the volumetric throughput during the clarification operation of the composition comprising
the biomacromolecule of interest. Data is shown for four separate MF experiments:
the runs shown by the open and filled triangles represent harvest streams that have
been pH 5.0 adjusted including the addition of 10mM CoCl2; and the runs shown by the open and filled circles represent unadjusted harvest streams
with no CoCl2 added. The data shows the MF retention coefficient is lower for pH adjusted harvest
feed containing 10mM CoCl2 for all loading ratios studied. The data indicates that runs containing the Co2+ divalent ions show complete recovery of the desired protein as compared to the 20%
yield loss with the runs using unadjusted harvest feed.
DETAILED DESCRIPTION OF THE INVENTION
[0010] The present invention is directed to a method as defined in the appendant claims.
The inventors of the present invention found that when attempting to separate (by
filtration) an industrial-scale amount of antibody from a bioreactor containing cell
culture with an initial high density of biological material, the transmembrane pressure
drop across the filter increased significantly, presumably as a result of fouling
of the membrane surface due to the high concentration of cellular material. Increased
filter fouling in turn adversely impacted quantity of the antibody recovered, lowered
clarification yields, and resulted in relatively higher impurity levels in the permeate
stream. In some cases, the transmembrane pressure increased to values beyond the mechanical
capabilities of the filter, thus causing the operation to stop before completion and
resulting in significantly lower product yield.
[0011] In order to reduce filter transmembrane pressure, increase protein recovery in the
permeate stream, and decrease the amount of impurities in the permeate stream, the
method of the present invention lowers the pH of the harvest feed before filtration,
causing flocculation of large cells and cellular debris along with precipitation of
other impurities (such as DNA). It was found that flocculation of impurities (cells,
cellular debris and DNA) into large particles improved mass transfer of the composition
near the surface of the filter, thus reducing transmembrane pressure across the filter
at a predetermined permeate stream flux or flowrate. In addition, co-precipitation
of impurities caused by dropping the harvest stream pH results in better retention
of these impurities on the filter, and thus a reduction of impurity levels in the
permeate stream.
[0012] Lowering the pH of the composition was found to induce precipitation of the desired
antibody in the harvest feed, resulting in retention of the desired antibody by the
membrane and reducing the amount of the desired antibody in the permeate stream. The
method of the present invention further provides that the addition of divalent cations
to pH-adjusted harvest feed prior to filtration selectively reduces the coprecipitation
of the antibody and increases the amount of antibody recovered in the permeate stream,
while not impacting the amount of impurities.
[0013] The methods of the present invention are useful for antibodies from impurities in
a composition.
[0014] It is to be noted, unless otherwise clear from the context, that the term "a" or
"an" entity refers to one or more of that entity; for example, "a protein," is understood
to represent one or more proteins. As such, the terms "a" (or "an"), "one or more,"
and "at least one" can be used interchangeably herein.
[0015] The terms "isolating" and "isolation" refer to separating a biomacromolecule, i.e.
an antibody, from at least one other undesired component or impurity found in the
composition. The term "isolating" includes "purifying" and "clarifying." No particular
level of isolation of a biomacromolecule, i.e. an antibody, is required, however in
some embodiments, at least 50%, 70%, 80%, 90%, or 95% (w/w) of an impurity is separated
from the biomacromolecule. For example, in some embodiments, isolation of a biomacromolecule,
i.e. an antibody, would comprise separating the biomacromolecule from 80% of the HCP
present originally in the composition.
[0016] The terms "clarifying" and "clarification" refer to the removal of large particles
from a composition. For example, as applied to cellular cultures and growth media,
the term "clarifying" refers to,
e.g., the removal of prokaryotic and eukaryotic
(e.g., mammalian) cells, lipids, and/or nucleic acids (e.g., chromosomal and plasmid DNA)
from the cell culture. In some embodiments, the method of the present invention comprises
(a) lowering the pH of the composition, allowing an impurity to flocculate within
the composition, (b) adding a divalent cation to the composition; and (c) separating
the biomacromolecule from an impurity in the composition. No particular level of flocculation
of an impurity is required, however in some embodiments, at least 50%, 70%, 80%, 90%,
or 95% (w/w) of an impurity is flocculated. For example, in some embodiments, clarification
of a biomacromolecule, i.e. an antibody, comprise flocculating 80% of the mammalian
cells present in a composition. Flocculation can be measured by methods known to those
in the art, including spectrophotographic methods such as a turbidimeter.
[0017] The terms "purifying" and "purification" refer to separating the biomacromolecule
of the invention from an impurity or other contaminants in the composition, regardless
of the size of the impurity. Thus, the term purification would encompass "clarification,"
but it would additionally encompass impurities smaller in size than those removed
during clarification,
e.
g., proteins, lipids, nucleic acids, and other forms of cellular debris, viral debris,
contaminating bacterial debris, media components, and the like. No particular level
of purification of a biomacromolecule, i.e. an antibody, is required, however in some
embodiments, at least 50%, 70%, 80%, 90%, or 95% (w/w) of an impurity is purified
from the biomacromolecule. For example, in some embodiments, purification of a biomacromolecule,
i.e. an antibody, would comprise separating the biomacromolecule from 80% of the HCP
present originally in the composition.
[0018] The terms "biological biomacromolecule" or "biomacromolecule" as used herein refer
to an antibody, e.g., a monoclonal antibody.
[0019] As used herein, the term "protein" is intended to encompass a singular "protein"
as well as plural "proteins." Thus, as used herein, terms including, but not limited
to "peptide," "polypeptide," "amino acid chain," or any other term used to refer to
a chain or chains of amino acids, are included in the definition of a "protein," and
the term "protein" may be used instead of, or interchangeably with, any of these terms.
The term further includes proteins which have undergone post-translational modifications,
for example, glycosylation, acetylation, phosphorylation, amidation, derivatization
by known protecting/blocking groups, proteolytic cleavage, or modification by non-naturally
occurring amino acids. Proteins also include polypeptides which form multimers,
e.
g., dimers, trimers, etc. The term protein also includes fusions proteins,
e.g., a protein that is produced via a gene fusion process in which a protein (or fragment
of a protein) is attached to an antibody (or fragment of antibody). Examples of fusion
proteins include disulfide-linked bifunctional proteins comprised of linked Fc regions
from human IgG1 and human IgE; and lymphotoxin beta receptor immunoglobulin G1.
[0020] Antibodies are purified according to the method of the present invention. The term
"antibody" refers to polyclonal, monoclonal, multispecific, human, humanized or chimeric
antibodies, single chain antibodies, Fab fragments, F(ab')2 fragments, fragments produced
by a Fab expression library, anti-idiotypic (anti-Id) antibodies (including,
e.g., anti-Id antibodies to antibodies of the invention), and epitope-binding fragments
of any of the above. In some embodiments, the term "antibody" refers to a monoclonal
antibody. The term "antibody" also refers to immunoglobulin molecules and immunologically
active portions of immunoglobulin molecules,
i.e., molecules that contain an antigen binding site that immunospecifically binds an antigen.
The immunoglobulin molecules that can be purified by the method of the invention can
be of any type
(e.g., IgG, IgE, IgM, IgD, IgA and IgY), class (e.g., IgG1, IgG2, IgG3, and IgG4) or subclass
of immunoglobulin molecule. Antibodies of the present invention also include chimeric,
single chain, and humanized antibodies. Examples of antibodies of the present invention
include commercialized antibodies, such as natalizmab (humanized anti-a4 integrin
monoclonal antibody), humanized Anti-Alpha V Beta 6 monoclonal antibody, humanized
anti-VLA1 IgG1 kappa monoclonal antibody; huB3F6 (humanized IgG1/kappa monoclonal
antibody).
[0021] Antibodies purified by the method of the invention may be from any animal origin
including birds and mammals. Preferably, the antibodies purified by the method of
the invention are human, murine (
e.g., mouse and rat), donkey, ship rabbit, goat, guinea pig, camel, horse, or chicken.
As used herein, "human" antibodies include antibodies having the amino acid sequence
of a human immunoglobulin and include antibodies isolated from human immunoglobulin
libraries or from animals transgenic for one or more human immunoglobulin and that
do not express endogenous immunoglobulins. See,
e.g., U.S. Pat. No. 5,939,598 by Kucherlapati
et al. In some embodiments, the antibody include, but are not limited to, IgG1, IgG2, IgG3,
and IgG4 antibodies, including commercialized antobodies, such as natalizumab TYSABRI®,
Elan Pahrmaceuticals, San Diego, CA).
[0022] Antibodies that can be purified by the method of the invention include,
e.g., native antibodies, intact monoclonal antibodies, polyclonal antibodies, multispecific
antibodies (
e.g., bispecific antibodies) formed from at least two intact antibodies, antibody fragments
(
e.g., antibody fragments that bind to and/or recognize one or more antigens), humanized
antibodies, human antibodies (
Jakobovits et al., Proc. Natl. Acad. Sci. USA 90:2551 (1993),
Jakobovits et al., Nature 362:255-258 (1993),
Bruggermann et al., Year in Immunol. 7:33 (1993);
U.S. Patent Nos. 5,591,669 and
5,545,807), antibodies and antibody fragments isolated from antibody phage libraries (
McCafferty et al., Nature 348:552-554 (1990);
Clackson et al., Nature 352:624-628 (1991);
Marks et al., J. Mol. Biol. 222:581-597 (1991);
Marks et al., Bio/Technology 10:779-783 (1992);
Waterhouse et al., Nucl. Acids Res. 21:2265-2266 (1993)). The antibodies purified by the method of the invention may be recombinantly fused
to a heterologous polypeptide at the N- or C-terminus or chemically conjugated (including
covalently and non-covalently conjugations) to polypeptides or other compositions.
For example, antibodies purified by the method of the present invention may be recombinantly
fused or conjugated to molecules useful as labels in detection assays and effector
molecules such as heterologous polypeptides, drugs, or toxins. See,
e.g., PCT publications WO 92/08495;
WO 91/14438;
WO 89/12624;
U.S. Pat. No. 5,314,995; and
EP 396,387.
[0023] In some embodiments, the biomacromolecule or composition of the present invention
is pharmaceutically acceptable. "Pharmaceutically acceptable" refers to a biomacromolecule,
i.e. an antibody, or composition that is, within the scope of sound medical judgment,
suitable for contact with the tissues of human beings and animals without excessive
toxicity or other complications commensurate with a reasonable benefit/risk ratio.
[0024] The term "composition" in the present invention refers to a mixture of one or more
molecules of the biomacromolecule of the present invention and optionally at least
one impurity, wherein the impurity and the biomacromolecule are not the same. In some
embodiments, the composition comprises the biomacromolecule of the invention, a cellular
host organism (
e.g., mammalian cells), and a growth media sufficient for propagating the host organism
and allowing expression of the biomacromolecule of the invention. The selection and
use of growth medium are known to those in the art. In some embodiments, the growth
media is a cell culture media. Cell culture media vary according to the type of cell
culture being propagated. In some embodiments, the cell culture media is a commercially
available media. In some embodiments, the composition comprises a growth media which
contains
e.g., inorganic salts, carbohydrates
(e.g., sugars such as glucose, galactose, maltose or fructose) amino acids, vitamins (
e.g., B group vitamins
(e.g., B12), vitamin A vitamin E, riboflavin, thiamine and biotin), fatty acids and lipids
(
e.
g., cholesterol and steroids), proteins and peptides (
e.g., albumin, transferrin, fibronectin and fetuin), serum
(e.g., compositions comprising albumins, growth factors and growth inhibitors, such as,
fetal bovine serum. newborn calf serum and horse serum), trace elements (
e.g., zinc, copper, selenium and tricarboxylic acid intermediates) and combinations thereof.
Examples of growth medias include, but are not limited to, basal media (
e.g., MEM, DMEM, GMEM), complex media (RPMI 1640, Iscoves DMEM, Leibovitz L-15, Leibovitz
L-15, TC 100), serum free media (
e.
g., CHO, Ham F10 and derivatives, Ham F12, DMEM/F12). Common buffers found in growth
media include PBS, Hanks BSS, Earles salts, DPBS, HBSS, EBSS. Media for culturing
mammalian cells are well known in the art and are available from,
e.g., Sigma-Aldrich Corporation (St. Louis, MO), HyClone (Logan, UT), Invitrogen Corporation
(Carlsbad, CA), Cambrex Corporation (E. Rutherford, NJ), JRH Biosciences (Lenexa,
KS), Irvine Scientific (Santa Ana, CA), and others. Other components found in growth
media can include ascorbate, citrate, cysteine/cystine, glutamine, folic acid, glutathione,
linoleic acid, linolenic acid, lipoic acid, oleic acid, palmitic acid, pyridoxal/pyridoxine,
riboflavin, selenium, thiamine, transferrin. One of skill in the art will recognize
that there are modifications to growth media which would fall within the scope of
this invention.
[0025] In some embodiments, the composition further comprises a harvest feed. The term "harvest
feed" refers to a media in which cells are present in immediately before harvesting,
or a media in which harvested cells are placed immediately after harvesting and into
which the cells are resuspended. A harvest feed can include any of the compositions
listed above for growth media, or other media suitable for resuspending the harvested
cells or cellular fractions. For example, in some embodiments, the harvest media may
contain water, a buffer, osmotic agents, anti-degradation agents, etc.
[0026] The term "impurity" refers to one or more components of the composition that is different
from the biomacromolecule of the present invention. In some embodiments, the impurity
can include an intact mammalian cell (
e.g., Chinese hamster ovary cells (CHO cells) or murine myeloma cells (NSO cells)), or
partial cells,
e.g., cellular debris. In some embodiments, the impurity comprises a protein (
e.g., soluble or insoluble proteins, or fragments of proteins, such as HCP), lipid
(e.g., cell wall material), nucleic acid
(e.g., chromosomal or extrachromosomal DNA), ribonucleic acid (t-RNA or mRNA), or combinations
thereof, or any other cellular debris that is different from the biomacromolecule
of interest. In some embodiments, the impurity can originate from the host organism
that produced or contained the biomacromolecule of the invention. For example, an
impurity could be a cellular component of a prokaryotic or eukaryotic cell
(e.g., cell wall, cellular proteins, DNA or RNA, etc.) that expressed a protein of interest.
In some embodiments, the impurity is not from the host organism, e.g., an impurity
could be from the cell culture media or growth media, a buffer, or a media additive.
The impurity as used herein can include a single undesired component, or a combination
of several undesired components.
[0027] The biomacromolecule of the present invention can be isolated from a cell culture
comprising growth media and various eukaryotic cells,
e.g., mammalian cells. The mammalian cells of the present invention, including the mammalian
cells that are used in the methods of the invention, are any mammalian cells that
are capable of growing in culture. Exemplary mammalian cells include, e.g., CHO cells
(including CHO-K1, CHO DUKX-B11, CHO DG44), VERO, BHK, HeLa, CV1 (including Cos; Cos-7),
MDCK, 293, 3T3, C127, myeloma cell lines (especially murine), PC12, HEK-293 cells
(including HEK-293T and HEK-293E), PER C6, Sp2/0, NS0 and W138 cells. Mammalian cells
derived from any of the foregoing cells may also be used.
[0028] The biomacromolecule of the present invention can be isolated from a cell culture
comprising growth media and various prokaryotic cells,
e.g., E. coli, Bacillus subtilis, Salmonella typhimurium and various species within the genera
Pseudomonas, e.g., P. aeruginosa, yeast cells,
e.g., Saccharomyces, Pichia, Hansenula, Kluyveramyces, Schizosaccharomyces, Schwanniomyces and
Yarrowia, insect cells,
e.g., Trichoplusia, Lipidotera, Spodoptera, Drosophila and
Sf9, or plant cells,
e.g., Arabidopsis. One of skill in the art can select an appropriate cell line depending on the biomacromolecule
of the invention
[0029] In the present invention, the pH of the composition is adjusted to a pH lower than
that of the harvest feed. Compositions of the present invention, e.g., those comprising
a harvest feed, generally have a pH of about 6.0 to about 8.0, about 6.5 to about
7.5 or about 6.8 to about 7.2 without adjustment. In some embodiments, the pH of the
composition is lowered to a pH within a range of about 4.0 to about 5.0, about 4.0
to about 4.7, about 4.3 to about 5.0, or about 4.7 to about 5.0. In some embodiments,
the pH of the composition is lowered to within a range of about 4.0 to about 4.7.
some biomacromolecules of the invention, a pH lower than 3.5 results in denaturation
or instability of the biomacromolecule of the invention, and thus is not desireable.
While not being bound by any theory, in some embodiments the lowering of harvest feed
pH flocculates one or more components of the composition, predominantly cells and
cellular debris. In some embodiments, host cell DNA is flocculated, enabling easier
and/or more efficient isolation of the biomacromolecule of the invention. In some
embodiments, host cell proteins are flocculated, enabling easier and/or more efficient
isolation of the biomacromolecule of the invention. In some embodiments wherein filtration
is used to isolate the biomacromolecule, the aggregated large particles reduce the
fouling of the pores of a filter, thus allowing for greater filtering efficiency,
lower transmembrane pressures, and higher throughput volumes.
[0030] The pH of the composition of the present invention can be adjusted by various means.
In some embodiments, the pH is lowered by addition of an acid to the composition.
Suitable acids include, but are not limited to, strong acids such as perchloric acid
(HCLO
4), hydroiodic acid (HI), hydrobromic acid (HBr), hydrochloric acid (HCl), nitric acid
(HNO
3), sulfuric acid (diprotic) (H
2SO
4), or weak acids such as acetic acid (CH
3COOH) (
e.
g., glacial acetic acid), citric acid (C
6H
8O
7), formic acid (HCOOH), hydrocyanic acid (HCN), hydrogen sulfate ion (HSO
4-), or combinations of any of the acids listed above. In some embodiments, the pH of
the composition can be adjusted by use of buffers, such as phosphate buffers (
e.g., sodium and potassium phosphates), bicarbonate buffers, citrate buffers, borate buffers,
acetate buffers, tromethamine buffers, HEPES buffers, and combinations thereof.
[0031] While not being bound by any theory, in some embodiments of the present invention,
the lowering of the pH aids in aggregating large impurity particles, thereby reducing
"fouling,"
i.e., plugging or filling the pores, of the filter. Increased filter fouling can adversely
impact recovery of a desired biomacromolecule in the permeate stream, resulting in
low clarification yields and relatively higher impurity levels in the permeate stream.
In some cases, the fouling of the filter increases transmembrane pressure. Fouling
of the filter may increase transmembrane pressure value beyond the mechanical capabilities
of the filter, thus causing the filter operation to be stopped before completion and
resulting in significantly lower product recovery. In some embodiments, the lowered
pH can increase the purity and quality of the biomacromolecule of the invention recovered.
[0032] In some embodiments, the lowering of the pH of the composition results in coprecipitation
of the biomacromolecule of the invention as well as the impurity, resulting in reduced
recovery of the biomacromolecule in the extracellular media. Inventors of the present
invention have found that addition of divalent cations to the pH-adjusted composition
is suitable for increasing the recovery of the biomacromolecule of the invention.
The term "increased recovery" refers to a comparison of method of the present invention
relative to an identical method of purifying but without the addition of divalent
cations. For example, if Method A is the method of the present invention (except it
does not comprise addition of divalent cations to the harvest feed) and yields 100
mg of the biomacromolecule of the invention, and Method B is identical to Method A
(except Method B comprises addition of divalent cations to the harvest feed) and yielded
110 mg of biomacromolecule, then it would be determined that Method B has an "increased
recovery" of 10%. In some embodiments, the method of the present invention increases
recovery of the biomacromolecule by greater than 3%, 4%, 5%, 6%, 7%, 8%, 9%, 10%,
15%, 20% or 25%. In some embodiments, the method of the present invention increases
recovery up to 10%, 15%, 20%, 25%, 30% or 50%.
[0033] In the present invention, a divalent cation is added to the composition. Various
divalent cations exist and are known to those in the art, and include,
e.
g., calcium cation (Ca
2+), magnesium cation (Mg
2+), copper cation (Cu
2+), cobalt cation (Co
2+), manganese cation (Mn
2+), nickel cation (Ni
2+), beryllium cation (Be
2+), strontium cation (Sr
2+), barium cation (Ba
2+), radium cation (Ra
2+), zinc cation (Zu
2+), cadmium cation (Cd
2+), silver cation (Ag
2+), palladium cation (Pd
2+), rhodium cation (Rh
2+), and combinations thereof. One of skill in the art will realize that the cation
can exist in salt form,
e.g., a calcium salt such as CaCl
2 can produce a calcium cation when placed in an aqueous solution. Thus, as used herein,
the phrase "adding a divalent cation" would encompass not only the addition of a cation
in its charged stated, but also the addition of a salt or other compound that would
produce a divalent cation upon introduction into the composition of the present invention.
According to the invention the divalent cation is Co
2+ or Ni
2+, or their salts skill in the art can easily and quickly test many divalent cations
to determine which
[0034] Various concentrations of divalent cations in the composition are suitable for use
in the present invention. On of skill in the art will recognize that various amounts
of divalent cations are normally present in small amounts in the harvest feed (endogenous
divalent cations), and that various amounts of divalent cations can be added to the
harvest feed in accordance with the present invention (exogenous divalent cations).
In some embodiments, the concentration of the divalent cations comprises both exogenous
and endogenous cations. However, for practical purposes, since the amount of endogenous
is relatively small compared to the amount of exogenous divalent cations, the concentration
of the divalent cations can be calculated by simply considering the exogenous divalent
cations. In some embodiments, the divalent cation is present at a concentration of
about 0.5 mM to about 200 mM, about 1.0 mM to about 100 mM, about 2 mM to about 50
mM, about 5 mM to about 15 mM, or about 2 mM to about 20 mM in the composition. In
some embodiments, the divalent cation is present at a concentration of about 10 mM
in the composition. While not being bound by any methodology, a suitable concentration
of divalent cation can be determined by method similar to that found in Example 14,
wherein various concentrations of divalent cation are added to a pH-adjusted composition
comprising a biomacromolecule, and then determining the lowest concentration at which
a maximum amount of biomacromolecule can be recovered. One of skill in the art will
understand that different concentrations of cations may be required for various biomacromolecules.
[0035] Various means can be used to separate the biomacromolecule of the present invention
from one or more impurities. Examples of means of separating the biomacromolecule
from an impurity include, without limitation, precipitation, immunoprecipitation,
chromatography, filtration, centrifugation, and combinations thereof. In some embodiments,
the separating of the biomacromolecule from the impurity is achieved by the use of
a filter. The term "filtration" or "filtering" refers to the process of removing suspended
particles from a composition by passing the composition through one or more semi-permeable
membranes (or medium) of a specified pore size diameter. The term "permeate stream"
when referring to filtration, refers to the fraction of the composition that passes
through the filter pores during filtration. The term "retentate stream" when referring
to filtration, refers to the fraction of the composition that remains on the filter
or that does not pass through the filter pores during filtration. In some embodiments,
after filtration the biomacromolecule of the present invention is substantially in
the permeate stream (
i. e., it passes through the filter pores and is collected), while an impurity (
e.g., cellular debris, DNA, and/or HCP) is substantially in the retentate stream. In some
embodiments, "bench scale" filtration can be used to predict appropriate conditions
for industrial scale filtration.
[0036] Suitable filter types, chemistries, and module configurations for purifying particular
biomacromolecules are known to those in the art and can be selected based on various
factors,
e.g., the amount and size of the components of the composition to be filtered, the volume
of the composition to be filtered, and the cell density and viability of the composition
to be filtered. In some embodiments, filters, such as membrane filters, plate filters,
cartridge filters, bag filters, pressure leaf filters, rotary drum filters or vacuum
filters can be used. In some embodiments, a depth filter or a cross filter is used.
The types of crossflow filter modules that apply in the present invention include
hollow fiber, tubular, flat plate (plate-and-frame), spiral wound, or vortex flow
(
e.g., rotating) filter geometries. In some embodiments, a tangential flow filter is used.
In some embodiments, hollow fibers, tubular, and flat-sheet membrane modules were
utilized in a tangential flow (cross-flow) mode. Commercially available filters that
can be employed are manufactured for various manufacturing vendors such as Millipore
Corporation (Billerica, MA), Pall Corporation (East Hills, NY), GE Healthcare Sciences
(Piscataway, NJ), and Sartorius Corporation (Goettingen, Germany).
[0037] The pore diameter in the filters of the present invention can vary according to the
type of biomacromolecule being isolated and the type of impurities present in the
composition. In some embodiments, the filter pore diameters can be 0.1µm to 1.0µm,
0.2µm to 0.8µm, or 0.2µm to 0.65µm in diameter.
[0038] Movement of a composition, such as a harvest feed, through a filter during filtration
generates a transmembrane pressure resulting from the membrane resistance. As the
membrane surface becomes accumulated (or polarized) with cellular material, there
is an increased resistance to flow across the membrane at a constant flowrate, thus
causing the driving force or transmembrane pressure to increase. If the amount of
cellular material near the surface of the membrane is reduced, or if the membrane
is less polarized, the transmembrane pressure tends to remain substantially constant.
Methods to calculate transmembrane potential are know to those in the art, and include
the use of pressure transducers or gauges. In some embodiments of the present invention,
the transmembrane pressure can be calculated by taking the difference between the
average of the feed and retentate stream outlet pressure and the permeate stream pressure.
Figure 8 show a schematic of the transmembrane pressure calculation.
[0039] Generally, during filtration of a composition,
e.g., a harvest feed, that has not been pH-adjusted, the transmembrane pressure of a filter
increases significantly as more of the composition is loaded onto the filter. For
example, in some embodiments the transmembrane pressure increases 5 psi, 7 psi, 10
psi, 15 psi or 20 psi or greater from the start of the filtration process (when the
first amount of the composition is placed in the filter) to the end of the filtration
process (typically following a 7-10x concentration of cellular material and a 3-5x
diafiltration) as the pores of the filter become clogged. For example, as can be seen
in FIG. 12 (Feed composition A) and FIG. 13 (Feed composition A), the transmembrane
pressure during filtration of a harvest feed solution in which the pH has not been
adjusted can increase from about less than 1 psi at the beginning of loading the harvest
feed onto the filter to about greater than 10 psi after loading 60 liters/m
2 of harvest feed onto a filter (a 1000% increase in transmembrane pressure). Thus,
the term "substantially constant" where referring to the transmembrane pressure, refers
to transmembrane pressures that do not increase greater than 4 psi, 3 psi or 2 psi
over the course of filtration. A substantially constant transmembrane pressure is
exemplified by the harvest feed compositions B through G in FIG. 12 and feed compositions
B through F in FIG. 13. Each of these compositions start at a transmembrane pressure
of about less than 1 psi and over the course of filtration, the transmembrane pressure
on the filters which are filtering these compositions do no exceed 2 psi. Thus, the
transmembrane pressures on the filters which are filtering compositions B through
G in FIG. 12 and feed compositions B through F in FIG. 13 are considered to be substantially
constant.
[0040] In some embodiments, the method of the present invention decreases protein filter
rejection. The term "protein filter rejection" can be exemplified by the equation
R = (1-[C
P/C
R]), wherein R represents protein filter rejection coefficient, C
P is the instantaneous permeate concentration of the biomacromolecule of interest,
and C
R is the instantaneous retentate concentration of the biomacromolecule of interest.
In some embodiments, the value of the protein filter rejection coefficient is lower
for a given volumetric throughput relative to a filter rejection coefficient is lower
for a given volumetric throughput relative to a biomacromolecule; i.e. an antibody
in a composition without pH adjustment and/or divalent cation addition. See for example
FIG. 18.
[0041] When isolating biomacromolecules, of the invention in some embodiments large volumes
of a composition (
e.
g., harvest feed) can be present, e.g., during commercial manufacturing processes.
Large volumes present several challenges for purification processes. For example,
the effect that a small change in flow rate through a filter has on the recovery of
the isolated biomacromolecule is amplified when large volumes are used. Likewise,
when using large volumes, the effect that an increase in cell density in a harvest
feed has on product recovery is also amplified. Thus, the use of large volumes of
a composition present unique problems that are amplified and have greater ramifications
relative to the use of smaller volumes. Thus, in some embodiments the present invention
is directed to a method of isolating an antibody present a biomacromolecule present
in a large volume of a composition. The term "large volume" refers to volumes associated
with the commercial and/or industrial production of a biomacromolecule of the invention.
In some embodiments to volumes associated with the commercial and/or industrial production
of a biomacromolecule of the invention. In some embodiments the term "large volume"
refers to 10 to 2000 liters, 20 to 1000 liters or 50 to 500 liters.
[0042] In some embodiments, it is beneficial or desirable to harvest the biomacromolecule
of the invention from a high cell density composition (
e.
g., harvest feed). High cell density compositions present unique problems relative
to normal cell density compositions. For example, high cell density compositions can
have higher amounts of impurities present in the composition, thereby increasing the
amount of impurities that need to be removed during the purification process. Thus,
a higher cell density composition can foul a filter more quickly, thereby prohibiting
filtration of the composition. In some embodiments, high cell density compositions
require the use of more filters, or filters with larger surface areas. Both of these
requirements can result in greater costs associated with filtration and/or loss of
product. In the present invention, the pH of the composition is lowered, thereby removing
some impurities, and allowing the purification of higher cell density compositions.
Thus, some embodiments in the present invention are directed to a method of isolating
a biomacromolecule, i.e. an antibody, some embodiments in the-present invention are
directed to a method of isolating a biomacromolecule; i.e. an antibody present in
a high cell density composition. The term "high cell density" generally refers to
cell densities in a harvest feed of about 1 x 10
5 to 3.5 x 10
7 , about 1.0 x 10
6 to about 1.0 x 10
7, or about 5.0 x 10
6 to about 9.0 x 10
6 cells per ml for mammalian cells. Of course, one of skill in the art will appreciate
that various cells traditionally grow at different cell densities. Thus, in some embodiments,
"high cell density" cell cultures refers to cell cultures containing cells at a density
higher than the density traditionally practiced for that cell line.
[0043] In some embodiments, the method of the present invention comprises separating the
biomacromolecule from an impurity by subjecting the composition to a centrifugal force
(
i.e., centrifugation), where centrifugation forms a supernatant and a precipitate. In some
embodiments, the centrifugation forms a supernatant substantially free of an impurity
(cells or cellular debris) and a concentrated cell/cellular debris precipitant. The
term "precipitate," when referring to centrifugation, refers to the fraction of the
composition that is precipitated (or pelleted) during centrifugation to form a cell/cell
debris mass. The term "supernatant," refers to the fraction of the composition that
is not precipitated (or pelleted) during centrifugation, for example, the fraction
of the composition that remains in an aqueous phase in the composition and is substantially
cell free. According to the method of the invention, after centrifugation the biomacromolecule
of the present invention is substantially in the supernatant (
i.e., it remains substantially suspended in the liquid fraction of the composition). In
some embodiments, a density gradient is used to separate the biomacromolecule from
the impurity. Thus, in some embodiments both the biomacromolecule and the impurity
remain in the supernatant after centrifugation, albeit at different densities and
thus different locations in the centrifugation apparatus.
[0044] Various centrifugation apparatuses can be used. In some embodiments, the centrifugation
can be accomplished by disc stack centrifugation. In some embodiments, "bench scale"
filtration can be used to predict appropriate conditions for industrial scale filtration.
Centrifugation variables can be varied to achieve optimal isolation of the biomacromolecule
of interest. For example, in some embodiments, various rotational speeds or flow rates
can be used to increase the quality of biomacromolecule recovery, and/or the quantity
of biomacromolecule recovery.
[0045] The steps of the method of the present invention can be ordered in various sequences.
For example, in some embodiments of the present invention, the lowering the pH and
the adding a divalent cation occurs before the separating the biomacromolecule away
from the impurity. In some embodiments, the pH of the harvest media is adjusted first,
the divalent cation is added, and then the biomacromolecule is separated from an impurity.
In other embodiments, the divalent cation is added first, the pH of the composition
is adjusted, and then the biomacromolecule is separated. In some embodiments, one
or more purification procedures may occur between either (a) the lowering of the pH,
(b) the adding of divalent cations, or (c) the separating the biomacromolecule away
from the impurity. Alternatively, the steps (a), (b), or (c) of the method of the
present invention can be contiguous, e.g., no additional purification procedures occur
between steps (a), (b) and (c). However, when steps (a), (b) and (c) are contiguous,
additional purification procedures can occur before or after steps (a), (b) and (c).
Examples
[0046] The invention will be described in greater detail by way of specific examples. The
following examples are offered for illustrative purposes, and are not intended to
limit the invention in any manner. Those of skill in the art will readily recognize
a variety of non-critical parameters that can be changed or modified to yield alternative
embodiments in accordance with the invention.
Example 1
[0047] The effects of pH on the flocculation of cells was investigated by comparing the
particle size distributions of cultured cells at various pH levels. Cell culture fluid
was harvested from two bioreactors, and then separated into three different groups,
one control group and two experimental groups. The experimental groups were adjusted
to approximately pH 4.5 or 5.75 using 25% acetic acid. The control group was left
unadjusted at approximately pH 7.1. The samples of each group were then analyzed using
the Beckman Coulter, Multisizer III (Fullerton, CA) and the average diameter was determined.
Results are presented in FIG. 1.
[0048] The experimental results indicate that at a pH less than 5.75, there is a pronounced
reduction of particles with a diameter less than 2.5 µm, and an increasing population
of particles 2.5 to 6.0 µm.
Example 2
[0049] The effects of pH changes on the flocculation of cells was investigated by staining
cells at various pH levels with Trypan Blue and analyzing them with a CEDEX cell analyzer
(Flonamics, Madison WI). As described in Example 1, cell culture fluid was harvested
from two bioreactors, and then separated into three different groups, one control
group and two experimental groups. The experimental groups were pH adjusted to approximately
4.5 and 5.75 using 25% acetic acid. The control group was left unadjusted at approximately
pH 7.1. TB stain was then added, and analyzed using magnified images. Results are
presented in FIG. 2.
[0050] TB dye stained dead cells by permeating through and staining intracellular proteins.
The data presented for cells at pH 4.50 shows many more stained particulates per unit
area relative to that of pH 7.00 and 5.75. This suggests that not only TB stained
dead cells, but also other protein aggregates of similar sizes are being stained.
This observation further supports the hypothesis that precipitation and flocculation
of proteins takes place at the lower pH.
Example 3
[0051] The effects of pH changes on supernatant turbidity was determined. Cell cultures
were divided into various aliquots, and then adjusted to a pH of 4.25, 4.5, 4.75,
5.0, 5.25, 5.5, 6.0 or unadjusted (pH 7.1). Settling velocity experiments were then
performed, and the turbidity of the resulting supernatant was measured. The results
are presented in FIG. 3.
[0052] The experimental results indicate the supernate turbidity increase with increase
in pH level. This suggests that at lower pH ranges, the cells are flocculating and
being removed during the settling velocity experiments.
Example 4
[0053] IgG1 and IgG4 class monoclonal antibodies were produced by recombinant Chinese Hamster
Ovary (CHO) or murine (NSO) cell fermentation. A vial of Master Cell Bank cells was
thawed and expanded via various shaker flasks and bioreactor sizes in the inoculum
preparation. The inoculum preparation and production bioreactor medium was made using
animal-free components. Temperature, pH and dissolved oxygen levels were controlled
in all bioreactor stated.
Example 5
[0054] The harvest feed of Example 4 was either (a) not pH-adjusted, or (b) adjusted to
a pH of 4.7. The adjusted (or unadjusted) harvest feed was then passed through a microfilter
in a tangential-flow filtration mode at a constant recirculation and permeate flowrate.
The transmembrane pressure during the filtration process was calculated (see FIG.
11) for both the pH-adjusted harvest feed and the non-pH adjusted harvest feed. Results
of the measurement of transmembrane pressure are provided for IgG1 class antibody
and IgG4 class antibody in FIG. 12 and FIG. 13, respectively. The figures demonstrate
that the lowering of a harvest feed to a pH ranging from 4.7 to 5.3 results in a low
and relatively constant transmembrane pressure when the harvest feed is applied to
a filter, whereas a harvest feed that is not pH adjusted does result in a sharp increase
in transmembrane pressure when applied to a filter.
[0055] While not being bound by a particular theory, the evidence suggests that the lowering
of the pH resulted in increased flocculation of cellular material, resulting in larger
particles which were less likely to interfere with the pores on the microfilter. The
clarification was performed in crossflow mode where the harvest feed was directed
into the microfiltration module containing microporus membranes, the retained cells
and cellular debris were retained and recirculated back into the feed vessel while
the cell-free permeate passed across the filter membrane and was collected (see FIG.
10 for schematic). The permeate was drawn from the system at a constant flow rate
and collected in a vessel. The separation by microfiltration of a desired protein
from a bioreactor containing cell culture with an initial high density of biological
material, resulted in an increased transmembrane pressure drop across the microfilter
as a result of fouling of membrane filter surface by high concentration of cellular
material. In order to reduce the filter transmembrane pressure drop to increase robustness
of the microfiltration operation and increase protein recovery in the permeate, the
pH of the harvest feed was lowered to 4.0-5.0 before microfiltration. The lowering
of the pH caused flocculation of large cells and cellular debris along with precipitation
of impurities (such as DNA). The flocculation of impurities (cells, cell debris and
DNA) into large particles improved mass transfer of the composition near the surface
of the filter thus reducing the pressure drop driving forces across the filter at
a predetermined permeate flux or flowrate (See FIGS. 3 and 4).
Example 6
[0056] The effect of lowering pH levels on removal of host cell proteins and DNA was investigated.
FIG. 5 demonstrations that lower pH levels increase both host cell protein and DNA
removal from a solution. Cell cultures were divided into different aliquots, then
the pH of the sample of each aliquot was adjusted. The samples were then centrifuged,
and the amount of host cell protein or DNA was determined. The data indicates that
host cell proteins were reduced by greater than 50% at a pH of less than 5.0, while
DNA was reduced by greater than 90% at a pH of less than 5.0. It is likely that these
impurities precipitate out of solution and are removed by the centrifugation.
Example 7
[0057] The effect of lowering pH on the quality and recovery of two monoclonal antibodies
was investigated. FIG. 6 indicates that protein recovery of both monoclonal antibodies
decreased below pH 6.0 by 5% to 10%, and decreased greatly for antibody A below 4.5.
This dramatic decrease in protein recovery was accompanied by increased levels of
aggregate and acidic variants, illustrated here by decreasing monomer and main isomform
levels, indicating molecular instability at low pH. Antibody B showed no significant
degradation over the pH range observed.
Example 8
[0058] The effect of lowering pH on the efficiency of a hollow fiber microfiltration (MF)
was investigated. FIG. 7 demonstrates that microfiltration harvest product quality
and impurity results agreed will the results of FIG. 5, with host cell proteins and
DNA being reduced by 50% and 90% respectively at pH less than 5.0. Clarified turbidity
levels were lower than initial experiments of FIG. 3 due to the filtration process
involved with the microfiltration unit operation. Still, the results indicate greater
than a 2-fold reduction in turbidity with cell culture conditioning to lower pH levels.
[0059] Bench scale MF yield results are illustrated in FIG. 8. Results illustrate yield
losses below pH 4.5 as seen during initial conditioning experiments. Additionally,
substantial losses were observed at pH 5.5 or higher due to membrane fouling during
MF processing. Flocculation effects provided by cell culture conditioning at lower
pH levels reduce this plugging.
Example 9
[0060] The effect of lowering pH on harvesting a monoclonal antibody was investigated. Antibody
B centrifugation results also agreed with the clarification results of FIG. 3 and
FIG. 5. Overall yield numbers for bench scale centrifuge runs were lower than initial
clarification experiments due to the additive product losses from conditioning and
the centrifugation unit operation. Also, turbidity levels were higher than clarification
experiments due to shear difference between bench and pilot scale centrifuges. However,
both of these process outputs follow the same trends seen during clarification experiments.
Increased variability in the centrifuge data points was due to varied centrifuge operating
conditions.
Example 10
[0061] The effect of lowering pH levels on the turbidity of the harvest feed of Example
4 was determined. The pH of the harvest feed containing either MgCl
2, CaCl
2, or NaCl was either (1) not adjusted (pH 6.9-7.2), (2) adjusted to a pH of 6.0, (3)
adjusted to a pH of 5.0, (4) adjusted to a pH of 5.0, (5) adjusted to a pH of 4.5,
or (6) adjusted to a pH of 4.0. The data was generated by pH adjusting aliquots of
harvest feed to the specified pH using 25% v/v acetic acid, allowing the flocculation
and settling of the cellular mass to occur for about 2 to 24 hrs and measuring the
turbidity (degree of clarity) of the clear supernatant using a turbidimeter. A general
reduction in supernatant turbidity occurred as the pH of the adjusted harvest material
dropped, indicating a clearer supernatant at lower pH values (see FIG. 4). This reduction
in turbidity was a result of a higher degree of cellular flocculation that occured
at lower pH levels, resulting in more rapid settling of cellular mass. The flocculation
and improved settling provided for more efficient micro filtration operational performance
(see FIG. 12 or FIG. 13).
Example 11
[0062] The effect of pH adjustment on the recovery of DNA contaminants from the harvest
feed was determined by collecting the harvest feed as described in Example 4 and then
either (a) not adjusting the pH of the harvest feed, or (b) lowering the pH of the
harvest feed to 4.7. Both the adjusted and the unadjusted harvest feed was then allowed
to settle for 24 hours. Samples were taken of the clear supernatant and tested for
DNA. The amount of DNA remaining in the cell free harvest was determined by Quantitative
Polymerase Chain Reaction (QPCR). FIG. 15 demonstrates that there is about a 1.5 to
3 log reduction in DNA contaminants in the harvest feed that has been adjusted to
a pH of 4.7 relative to a harvest feed that has not been pH adjusted.
Example 12
[0063] The effect of pH adjustment on the recovery of a monoclonal antibody (IgG4) and a
fusion protein was determined by collecting the harvest feed as described in Example
4. The harvest feed was then adjusted to a pH of either 7.0, 5.0 or 4.0 and allowed
to settle for 2-24 hours. Samples were taken of the clear supernatant and tested for
protein titer. The amount of antibody or fusion protein present in the cell-free harvest
feed was then determined by Protein G titer assay. FIG. 14 shows the effects of pH
change on the product protein titer in the conditioned Harvest stream. Data is shown
for an IgG4 class antibody and a fusion protein. The data for both types of proteins
demonstrate that there was significant reduction in protein as the pH drops from 7.0
to 4.0. The amount of protein loss due to pH induced precipitation ranged from 3%
to 52%. The method of determining protein concentration of the composition was based
on affinity capture of the antibody or fusion protein on Protein G coupled HPLC resin.
Example 13
[0064] The effect of addition of divalent cations to pH-adjusted harvest feed on protein
was investigated by adding 10 mM of various divalent cations to the harvest feed,
and then lowering the pH of the harvest feed to 5.0. The resulting harvest feed was
allowed to settle for 2-24 hours. Samples were taken of the clear supernatant and
tested for protein titer. The titer was compared to a control experiment (pH 5.0 adjusted
harvest with no divalent cations). The amount of antibody or fusion protein present
in the cell-free harvest feed was determined by Protein G titer assay. FIG. 16 demonstrates
that addition of divalent cations generally reduced the loss of soluble antibody compared
to the control run without any additional ions. The y-axis represent a normalized
titer, which is represented by (titer in filtered harvest feed that has been treated
with cation) divided by (titer in filtered harvest feed that has not been treated
with cation).
Example 14
[0065] The effect of various concentrations of divalent cations added to pH-adjusted harvest
feed was investigated. Specifically, CoCl
2 (or Co
2+) was added to a final concentration of 1 mM, 2 mM, 5 mM, 10 mM or 20 mM in pH-adjusted
harvest feed. The amount of product recovered from the pH-adjusted harvest feed was
then assayed using Protein G to determine the effect of the various concentrations
of CoCl
2 on product recovery. FIG. 17 demonstrates that about 10 mM of CoCl
2 was sufficient to minimize product loss (or maximize product yield recovery) after
pH adjustment.
Example 15
[0066] The effects of the method of the present invention on protein filter rejection was
investigated as a function of cumulative volume processed through a microporous membrane
throughout the concentration and diafiltration phases. The protein filter rejection
coefficient was determined by measuring the retentate and permeate titers at various
time intervals and calculating the protein filter rejection coefficient with the equation
R=(1-[C
P/C
R]). Data is shown for four separate MF experiments: the runs shown by the open and
filled triangles represent harvest streams that have been pH 5.0 adjusted including
the addition of 10mM CoCl2; and runs shown by the open and filled circles represent
unadjusted harvest streams with no presence of CoCl2. The data shows the MF retention
coefficient is lower for pH adjusted harvest feed containing 10mM CoCl2 for all loading
ratios studied. For example, at typical large-scale loading ratios of 60-70L/m2, the
calculated rejection coefficients are ∼30% compared to nearly complete rejection (90-100%)
for the runs that were not pH adjusted. Also shown in Figure 6 are overall protein
recoveries from each microfiltration run. The runs containing the Co
2+ divalent ions show complete recovery of the desired protein as compared to the 20%
yield loss with the runs using unadjusted harvest feed
[0067] The effects of the method of the present invention on protein filter rejection was
investigated as a function of cumulative volume processed through a microporous membrane
throughout the concentration and diafiltration phases. The protein filter rejection
coefficient was determined by measuring the retentate and permeate titers at various
time intervals and calculating the protein filter rejection coefficient with the equation
R=(1-[C
P/C
R]). The data presented in FIG. 18 is from four separate microfiltration experiments:
in C and D, 10mM CoCl
2 was added to the harvest stream during pH adjustment; in A and B, no CoCl
2 was added. All harvest feeds were lowered to a pH of 5.0. FIG. 18 demonstrates that
the microfiltration retention coefficient was lower for harvest feeds adjusted with
10mM CoCl
2 for all loading ratios studied. For example, at typical large-scale loading ratios
of 60-70L/m
2, the calculated rejection coefficients for microfiltration runs C and D were ∼30%
compared to nearly complete rejection (90-100%) for the runs A and B. FIG. 18 also
demonstrates the overall protein recovery from each microfiltration run. The runs
containing the Co
2+ divalent ions (C and D) showed close to complete recovery of the desired protein,
whereas runs in which no Co
2+ was added (A and B) resulted in a 20% yield loss.